Re: [PATCH] ARM: tegra: cardhu: add dt entry for PMIC TPS65911.

[Laxman Dewangan <ldewangan-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>]

On Friday 06 July 2012 12:46 AM, Stephen Warren wrote:
> On 07/04/2012 09:07 AM, Laxman Dewangan wrote:
>> Tegra30 based platform "cardhu" have the power management
>> IC TPS65911 for the regulator.
>> Adding DT entry for this device.
>>
>> Signed-off-by: Laxman Dewangan<ldewangan-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>
> I recall there were some differences between Cardhu A02 and A04. Does
> this patch apply equally to both?

Yes, this piece is common. I already sent the splitting A02 and A04 
which is having some differences in the fixed regulator GPIOs. It is 
pending for internal review.

> Please have a look at the recent threads re: various Tegra20 boards'
> regulator .dts files, and see the results at:
>
> git://nv-tegra.nvidia.com/user/swarren/linux-2.6 linux-next_common
>
> Most of these comments are driven by comments made in the review of
> those patches.
>
> In the patch description, can you please specify where you got all the
> values from, and any discrepancies.

Ok, I will do.

>> diff --git a/arch/arm/boot/dts/tegra30-cardhu.dts b/arch/arm/boot/dts/tegra30-cardhu.dts
>> +				vdd1_reg: regulator@0 {
>> +					reg =<0>;
>> +					regulator-compatible = "vdd1";
>> +					regulator-name = "vdd_1v2_gen";
> It'd be good to list all the signals that these regulators drive
> directly; the schematics "rename" the signals quite a lot.

Will add wherever it is possible.

>> +					regulator-min-microvolt =<  600000>;
>> +					regulator-max-microvolt =<1500000>;
> Similarly, the contraints should list the exact voltage that's required
> of the regulators, not a range. Right now, there is no DVFS in the
> mainline kernel, so even for rails where DVFS could be used in the
> future, we should specify the single voltage we want these rails to run
> at without DVFS for now.
>
Fine, will provide the one voltage.
>> +					regulator-always-on;
>> +					regulator-boot-on;
> These properties aren't in the same order in all the nodes. It'd be nice
> to have them ordered the same way everywhere.
>
> I'm not sure if it really makes sense to specify regulator-boot-on,
> since there's no way to know what SW has run before the kernel which
> might have turned off the regulator; regulator-always-on seems to cover
> all necessary use-cases.